Industry

Customer Data Platform

Company

Segment

Data observability

Designing a Data Observability Tool for Enterprise Trust

Enterprise customers at Segment faced significant challenges detecting and resolving data pipeline issues. Without clear implementation feedback or debugging tools, teams relied heavily on Solution Architects — leading to delays, inefficiencies, and frustration that eroded trust. I led the design of a data observability tool that gave teams real-time visibility, intuitive debugging workflows, and confidence in their data pipeline health.

Role

Senior Product Designer (Strategy & UX)

Time

~6 months

Team

2 designers (staff designer, me), 1 researcher, customer success partners

Problem Space

  • No real-time visibility: Customers couldn’t easily see pipeline health.

  • Fragmented debugging: Required navigating multiple disconnected tools and pages.

  • Weak feedback loops: Limited monitoring features slowed implementations and troubleshooting.

  • Business risk: High frustration led to increased support costs, delayed time-to-value, and churn risk.


Research & Insights

  • Conducted 28 in-depth interviews with both technical (data engineers) and non-technical (analysts, managers) enterprise users.

  • Mapped pain points across the implementation and maintenance journey.

  • Created a Customer Journey Map (CJM) that highlighted friction points and opportunities.

  • Facilitated cross-functional workshop to align PM, engineering, and solution architects around top challenges.

Key Insight: Users didn’t just want alerts — they wanted clear, actionable context to resolve issues quickly without relying on specialized support.



Design Goals

  • Visibility: Provide a single place to monitor pipeline health.

  • Actionability: Give clear metrics and visual cues for debugging.

  • Efficiency: Streamline troubleshooting workflows around key JTBDs.

  • Trust: Reinforce confidence in Segment’s reliability for mission-critical data.


Process & Approach

  • Vision & Prioritization: Developed a 3-year vision for a unified observability UI covering Source → Destination. Backcasted into a Minimum Lovable Product (MLP).

  • Design & Iteration: Built and tested Figma prototypes focusing on:

    • Intuitive visualizations of pipeline health.

    • Error diagnostics with contextual guidance.

    • Simplified workflows for common debugging tasks.

  • Collaboration: Used CJM + prototypes to align stakeholders and influence both product roadmap and engineering architecture decisions.


Solution

The Observability Dashboard included:

  • Unified health overview for Sources, Destinations, and Transformations.

  • Real-time metrics (throughput, error rates, dropped events).

  • Contextual error reporting with links to affected configurations.

  • Streamlined debugging workflows, reducing the need for external tools or log-diving.


Impact / Results

  • Improved User-Centricity: Shifted the product team’s mindset from assumptions to user-driven insights, thanks to research and journey mapping.

  • Efficiency Gains: Reduced time spent on implementations and support by empowering customers with clearer debugging tools.

  • Strategic Alignment: Provided a roadmap and vision that connected near-term wins with long-term observability strategy, unifying product and engineering teams.

  • Customer Value: Enhanced pipeline health monitoring reduced friction, rebuilt trust, and addressed key churn drivers for enterprise accounts.


Reflection / Learnings

  • In enterprise contexts, trust is a design outcome as much as usability.

  • Creating a 3-year vision + MLP backcast helped balance immediate impact with strategic alignment.

  • Cross-functional artifacts (like CJMs) proved invaluable to unify diverse teams and influence system-level decisions.

  • Future opportunity: expand observability into predictive insights (e.g., anomaly detection, proactive recommendations).

Problem Space

  • No real-time visibility: Customers couldn’t easily see pipeline health.

  • Fragmented debugging: Required navigating multiple disconnected tools and pages.

  • Weak feedback loops: Limited monitoring features slowed implementations and troubleshooting.

  • Business risk: High frustration led to increased support costs, delayed time-to-value, and churn risk.


Research & Insights

  • Conducted 28 in-depth interviews with both technical (data engineers) and non-technical (analysts, managers) enterprise users.

  • Mapped pain points across the implementation and maintenance journey.

  • Created a Customer Journey Map (CJM) that highlighted friction points and opportunities.

  • Facilitated cross-functional workshop to align PM, engineering, and solution architects around top challenges.

Key Insight: Users didn’t just want alerts — they wanted clear, actionable context to resolve issues quickly without relying on specialized support.



Design Goals

  • Visibility: Provide a single place to monitor pipeline health.

  • Actionability: Give clear metrics and visual cues for debugging.

  • Efficiency: Streamline troubleshooting workflows around key JTBDs.

  • Trust: Reinforce confidence in Segment’s reliability for mission-critical data.


Process & Approach

  • Vision & Prioritization: Developed a 3-year vision for a unified observability UI covering Source → Destination. Backcasted into a Minimum Lovable Product (MLP).

  • Design & Iteration: Built and tested Figma prototypes focusing on:

    • Intuitive visualizations of pipeline health.

    • Error diagnostics with contextual guidance.

    • Simplified workflows for common debugging tasks.

  • Collaboration: Used CJM + prototypes to align stakeholders and influence both product roadmap and engineering architecture decisions.


Solution

The Observability Dashboard included:

  • Unified health overview for Sources, Destinations, and Transformations.

  • Real-time metrics (throughput, error rates, dropped events).

  • Contextual error reporting with links to affected configurations.

  • Streamlined debugging workflows, reducing the need for external tools or log-diving.


Impact / Results

  • Improved User-Centricity: Shifted the product team’s mindset from assumptions to user-driven insights, thanks to research and journey mapping.

  • Efficiency Gains: Reduced time spent on implementations and support by empowering customers with clearer debugging tools.

  • Strategic Alignment: Provided a roadmap and vision that connected near-term wins with long-term observability strategy, unifying product and engineering teams.

  • Customer Value: Enhanced pipeline health monitoring reduced friction, rebuilt trust, and addressed key churn drivers for enterprise accounts.


Reflection / Learnings

  • In enterprise contexts, trust is a design outcome as much as usability.

  • Creating a 3-year vision + MLP backcast helped balance immediate impact with strategic alignment.

  • Cross-functional artifacts (like CJMs) proved invaluable to unify diverse teams and influence system-level decisions.

  • Future opportunity: expand observability into predictive insights (e.g., anomaly detection, proactive recommendations).

Problem Space

  • No real-time visibility: Customers couldn’t easily see pipeline health.

  • Fragmented debugging: Required navigating multiple disconnected tools and pages.

  • Weak feedback loops: Limited monitoring features slowed implementations and troubleshooting.

  • Business risk: High frustration led to increased support costs, delayed time-to-value, and churn risk.


Research & Insights

  • Conducted 28 in-depth interviews with both technical (data engineers) and non-technical (analysts, managers) enterprise users.

  • Mapped pain points across the implementation and maintenance journey.

  • Created a Customer Journey Map (CJM) that highlighted friction points and opportunities.

  • Facilitated cross-functional workshop to align PM, engineering, and solution architects around top challenges.

Key Insight: Users didn’t just want alerts — they wanted clear, actionable context to resolve issues quickly without relying on specialized support.



Design Goals

  • Visibility: Provide a single place to monitor pipeline health.

  • Actionability: Give clear metrics and visual cues for debugging.

  • Efficiency: Streamline troubleshooting workflows around key JTBDs.

  • Trust: Reinforce confidence in Segment’s reliability for mission-critical data.


Process & Approach

  • Vision & Prioritization: Developed a 3-year vision for a unified observability UI covering Source → Destination. Backcasted into a Minimum Lovable Product (MLP).

  • Design & Iteration: Built and tested Figma prototypes focusing on:

    • Intuitive visualizations of pipeline health.

    • Error diagnostics with contextual guidance.

    • Simplified workflows for common debugging tasks.

  • Collaboration: Used CJM + prototypes to align stakeholders and influence both product roadmap and engineering architecture decisions.


Solution

The Observability Dashboard included:

  • Unified health overview for Sources, Destinations, and Transformations.

  • Real-time metrics (throughput, error rates, dropped events).

  • Contextual error reporting with links to affected configurations.

  • Streamlined debugging workflows, reducing the need for external tools or log-diving.


Impact / Results

  • Improved User-Centricity: Shifted the product team’s mindset from assumptions to user-driven insights, thanks to research and journey mapping.

  • Efficiency Gains: Reduced time spent on implementations and support by empowering customers with clearer debugging tools.

  • Strategic Alignment: Provided a roadmap and vision that connected near-term wins with long-term observability strategy, unifying product and engineering teams.

  • Customer Value: Enhanced pipeline health monitoring reduced friction, rebuilt trust, and addressed key churn drivers for enterprise accounts.


Reflection / Learnings

  • In enterprise contexts, trust is a design outcome as much as usability.

  • Creating a 3-year vision + MLP backcast helped balance immediate impact with strategic alignment.

  • Cross-functional artifacts (like CJMs) proved invaluable to unify diverse teams and influence system-level decisions.

  • Future opportunity: expand observability into predictive insights (e.g., anomaly detection, proactive recommendations).